NICEI - Northern Ireland Composite Economic Index
2023-04-19
Contents:
This document will outline the steps taken, including the code, to achieve the plots in the “Growth Rates” panel of the NICEI Q4 2022 factsheet which can be found here. No other panels have been completed but can be ameneded to this document if so desired.
Panel 1: Growth Rates
1. NICEI Trend to Q4 2022
Step 1: Installing relevant libraries
library(readxl)
library(janitor)
library(ggplot2)
library(dplyr)
library(reshape2)
library(tidyverse)Step 2: Data Import and Preparation
Loading data
Import from Excel spreadsheet in local directory
df <- read_excel("NICEI-Tables-Q4-2022.xlsx", sheet = "Table 1", skip = 1)Data Preparation
Selection
Dropping last 3 columns that are unused and merging the Year and Quarter column
df1 <- df %>%
select(c(0:5)) %>%
mutate(date = paste(Year, Quarter, sep="\nQ"), .before = 1, .keep = "unused")Restructuring
Now for restructuring. This code takes the data from the 4 columns and restructures it to 3 columns which contain the date, the variable (previous column headers i.e the variables of interest) and their respective values. This is useful for colouring plots based on category as is needed in this case.
df2<-melt(df1, id.vars="date") Step 3. Plotting
Now to plot the data to find out what it looks like and what more needs done
Initial plot
ggnicei<-ggplot() +
geom_line(data = df2,
aes(x=date, y=value,
group = variable,
color = variable)
)
ggnicei
The plot lines look correct but work needs done to make the legend legible and prettify the plot so it resembles the original plot published. To alter the X-axis values so they’re readable, a function will be created to return a function as the breaks argument; this, in turn, will return the appropriate x-axis spacing i.e the nth value specified when the function is called.
Plot formatting/aesthetics
#Create function
every_nth = function(n) {
return(
function(x){
x[c(TRUE, rep(FALSE, n - 1))]
}
)
}The plot can now be redone implementing some formatting such as:
- scale_colour_manual: this formats the plot in relation to colour
- geom_hline: this inserts a horizontal line
- annotate: this permits annotation on the plot
- theme: this has many possible components that all will affect the aesthetic of the plot
final_plot<-ggnicei +
scale_color_manual(values= c("black", "blue", "#cccc00"), # selecting colours for the lines
labels=c("NICEI", # creating the line labels for the legend
"Private Sector Component Index",
"Public Sector Component Index")) +
geom_hline(yintercept=100, linetype = 'dashed') + # generating the dashed horizontal line (baseline)
annotate("text", # Forming the text for the baseline
x="2014\nQ1",
y= 99,
label = "Baseline 2019 = 100",
size =3,
fontface = "bold") +
theme(line = element_blank(), # plot theme
panel.grid.major = element_blank(), # no major gridlines
panel.grid.minor = element_blank(), # no minor gridlines
panel.background = element_blank(), # blank background
plot.title = element_text(colour = "chartreuse3", # title formatting: colour and bold
face = "bold"),
axis.text = element_text(face = "bold"), # formatting axis text: bold
axis.line = element_line(colour = "NA"), # no axis lines
legend.title=element_blank(), # remove legend title
legend.key=element_blank(), # remove background colour from legend key
legend.position = c(0.15,0.15), # legend position
plot.margin = unit(c(2,2,2,2), "cm")) + # adjusting plot padding for exported image
scale_x_discrete(breaks = every_nth(n = 8)) + # implementing the function for axis labels every 8th (Q1 every other year)
scale_y_continuous(breaks = seq(75, 110, by=5), # formatting Y-axis: 75-115 in increments of 5
limits=c(75,110))+
annotate(geom = "point", # inserting red point at Y-value 105.7
x = "2022\nQ4",
y = 105.7,
colour = "red",
size = 2) +
annotate(geom = "label", # inserting blue label with white font stating 105.7
x = "2021\nQ4",
y = 108,
label = "105.7",
hjust = "left",
fill="blue",
color = "white") +
annotate("segment", # connecting line between point and label
x = "2022\nQ1",
xend = "2022\nQ4",
y = 107.8, yend = 105.7,
linewidth = 0.2) +
labs(x = "", # no X-axis title
y="", # no Y-axis title
title = "NICEI Trend to Q4 2022") # plot title
final_plot
This plot seems much more comparable to the original in the publication; however, the plot in the original pdf seems to have a thicker NICEI line. If this is desired, all that is required is to overlay the final_plot with the same plot line but to the required thickness.
df4<-filter(df2, variable == "NICEI") # extract the values associated with the required plot line
final_plot2<- final_plot + # add these extracted values to the previous plot, formatting appropriately
geom_line(data=df4,
aes(x=date,
y=value,
group = 1),
linewidth = 1)
final_plot2
Step 4. Export plot image
Instead of exporting the plot as an image manually within R Studio, a scripted export of plot image can be done; however, it must be noted that there is a decrease in text readability using this method and so will need increased. An example of the manually exported plot is in the local directory named, “NICEI_r_plot.png”. Scripted export can be done as follows:
png("NICEI-plot.png",
width = 950,
height = 750)
final_plot2
dev.off()2. Growth Rates by Index
Step 1: Installing relevant libraries
library(readxl)
library(xts)
library(formattable)
library(flextable)
library(tidyverse)
library(gridExtra)
library(grid)
library(officer)Step 2: Data Import, Cleaning and Preparation
Loading data
Read in data (excel format) from local directory
df <- read_excel("NICEI-Tables-Q4-2022.xlsx", sheet = "Table 1", skip = 1)
formattable(head(df))| Year | Quarter | NICEI | Private Sector | Public Sector | Services | Production | Construction |
|---|---|---|---|---|---|---|---|
| 2006 | 1 | 102.6 | 101.8 | 105.9 | 99.3 | 101.7 | 134.9 |
| 2006 | 2 | 103.0 | 102.5 | 105.9 | 99.9 | 103.3 | 134.8 |
| 2006 | 3 | 104.2 | 103.9 | 106.4 | 101.0 | 103.2 | 143.1 |
| 2006 | 4 | 104.6 | 104.5 | 106.0 | 102.0 | 103.1 | 142.9 |
| 2007 | 1 | 104.8 | 104.9 | 105.8 | 102.5 | 104.1 | 137.9 |
| 2007 | 2 | 105.5 | 105.8 | 106.0 | 103.2 | 104.7 | 143.3 |
Calculations
This table contains data not present in the raw data table, instead, it must be calculated. In this section each row of the final growth table will calculated before being appended to each other, creating the final table.
Calulating value differences
This code chunk below is creating the first 3 rows of the table - calculating the differences in the required data
Q_Q <- (df[nrow(df),] - df[nrow(df)-1,])/df[nrow(df)-1,] # Calculate Q/Q row
Y_Y <- (df[nrow(df),] - df[nrow(df)-4,])/df[nrow(df)-4,] # Calculate Y/Y row
Tri <- (df[nrow(df),] - df[nrow(df)-12,])/df[nrow(df)-12,] # Calculate Triennial rowRolling 4Q Average
The last row of the final table is the rolling 4Q average of the data. This requires more detailed work. Firstly the required data from the raw form must be selected
# Create new column with year and Quarter combined, separated with a space, and delete the two individual columns
df1<- df %>% mutate(date = paste(Year, Quarter, sep="\nQ"), .before = 1, .keep = "unused")Once selected, this must be converted to a time series object in order to apply the time series functions
# Convert from data frame to a time series object to enable use of moving/rolling average calculations
TS <- df1 %>% ts(frequency = 4, start = c(2006,1), end = c(2022, 4)) # frequency = 4 means data consists of quarterly samples.Now the required rolling 4Q averages can be returned
# Calculate the 4Q rolling average of dataset
r4Q <-TS %>% # all data (window() would be used if calculations were to be on data subsets)
as.xts() %>% # coerces the timeseries object to an eXtensible format required for time-based analysis
apply.yearly(mean) %>% # apply mean function to each distinct year of data
as.data.frame() # pipe data to data frame for additional work
# calculation of fractional difference for rolling 4Q annual average
r4q<-(r4Q[nrow(r4Q),] - r4Q[nrow(r4Q)-1,])/r4Q[nrow(r4Q)-1,]Step 3. Tabulation and aesthetics
Now all required rows have been obtained, it is time to combine and format accordingly as per orginial publication
Tabulation of individual rows
The following code chunk combines the rows and appends different column names which fit the formatting required for the arrow inclusion code (compose package)
dt<-rbind(Q_Q[,-c(1:2)],
Y_Y[,-c(1:2)],
Tri[,-c(1:2)],
r4q[,-1]) %>%
round(digits = 3)
# create the column of names
growth <- data.frame("Growth Rates" = c("Q/Q",
"Y/Y",
"Triennial \n(3Yr) Change",
"Rolling 4Q \nAnnual Ave")
)
DT<-cbind(growth, dt) # combine base table with names
colnames(DT)<-c("growth", "NICEI", "private", "public", "services", "production", "construction") #convert column names to single value names (needed for the compose later)
formattable(DT)| growth | NICEI | private | public | services | production | construction | |
|---|---|---|---|---|---|---|---|
| 1 | Q/Q | 0.014 | 0.015 | 0.007 | 0.010 | -0.006 | 0.088 |
| 2 | Y/Y | 0.013 | 0.010 | 0.023 | 0.018 | -0.011 | -0.007 |
| 3 | Triennial (3Yr) Change | 0.057 | 0.057 | 0.055 | 0.062 | 0.046 | 0.046 |
| 2022 Q4 | Rolling 4Q Annual Ave | 0.030 | 0.031 | 0.026 | 0.044 | 0.025 | -0.036 |
The values in the table are almost correct; the original publication has the values as percentages.
pDT<- DT %>% mutate(across(-c(1),percent)) #convert values to percentage using the formattable package
formattable(pDT)| growth | NICEI | private | public | services | production | construction | |
|---|---|---|---|---|---|---|---|
| 1 | Q/Q | 1.40% | 1.50% | 0.70% | 1.00% | -0.60% | 8.80% |
| 2 | Y/Y | 1.30% | 1.00% | 2.30% | 1.80% | -1.10% | -0.70% |
| 3 | Triennial (3Yr) Change | 5.70% | 5.70% | 5.50% | 6.20% | 4.60% | 4.60% |
| 2022 Q4 | Rolling 4Q Annual Ave | 3.00% | 3.10% | 2.60% | 4.40% | 2.50% | -3.60% |
Aesthetics
Now the table is complete in its basic form, the colouring and arrows need to be added The formattable package has been used to this point but moving forward the flextable package will be used instead as this permits more advanced formatting such as the addition of icons (the arrows in this case).
dft<-flextable(pDT) %>% theme_box() # pretty print of table using flextable (easier to customise than formattable)
#prettify table
dft<-bold(dft, part = "header") # bold header
dft<-bg(dft, bg = "olivedrab3", part = "header") # fill the header with colour
dfxt <- dft # rename for next step of adding arrows
dfxtgrowth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q | 1.40% | 1.50% | 0.70% | 1.00% | -0.60% | 8.80% |
Y/Y | 1.30% | 1.00% | 2.30% | 1.80% | -1.10% | -0.70% |
Triennial | 5.70% | 5.70% | 5.50% | 6.20% | 4.60% | 4.60% |
Rolling 4Q | 3.00% | 3.10% | 2.60% | 4.40% | 2.50% | -3.60% |
In the complex process of adding coloured arrows corresponding to values, each column must be addressed individually and the green arrows will be added first as this is the predominant colour. Then each column with red arrows must will be amended to reflect this. This is a manual process because there is not a consistent rule in the values upon which to allocate the arrows. In the final table, each value corresponds to its previous value in a different year/quarter, which is dependent on the category it is describing. There are R emoticon/graphics packages available but it was easier to find .png files and save them to the local directory and upload
dfxt2 <- flextable::compose(dfxt, #data
j = 2, #2nd column
value = as_paragraph( # allows concatenation of text chunks and images etc
as_image(src = "arrow_up.png", width = .15, height = .15), # arrow first as per the publication being replicated
" ", # space follows the arrow
as_chunk(NICEI, props = fp_text(color = "black") # The text to be added, which is the value in the column NICEI.
) # This line is why the column names needed changed earlier (line 49)
),
part = "body") # noting that the compose function is applied to the table body
dfxt2 #table checkgrowth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
| 1.50% | 0.70% | 1.00% | -0.60% | 8.80% |
Y/Y |
| 1.00% | 2.30% | 1.80% | -1.10% | -0.70% |
Triennial |
| 5.70% | 5.50% | 6.20% | 4.60% | 4.60% |
Rolling 4Q |
| 3.10% | 2.60% | 4.40% | 2.50% | -3.60% |
dfxt2 <- flextable::compose(dfxt2,
j = 3,
value = as_paragraph(
as_image(src = "arrow_up.png", width = .15, height = .15), " ",
as_chunk(private, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
| 0.70% | 1.00% | -0.60% | 8.80% |
Y/Y |
|
| 2.30% | 1.80% | -1.10% | -0.70% |
Triennial |
|
| 5.50% | 6.20% | 4.60% | 4.60% |
Rolling 4Q |
|
| 2.60% | 4.40% | 2.50% | -3.60% |
dfxt2 <- flextable::compose(dfxt2,
j = 4,
value = as_paragraph(
as_image(src = "arrow_up.png", width = .15, height = .15), " ",
as_chunk(public, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
| 1.00% | -0.60% | 8.80% |
Y/Y |
|
|
| 1.80% | -1.10% | -0.70% |
Triennial |
|
|
| 6.20% | 4.60% | 4.60% |
Rolling 4Q |
|
|
| 4.40% | 2.50% | -3.60% |
dfxt2 <- flextable::compose(dfxt2,
j = 5,
value = as_paragraph(
as_image(src = "arrow_up.png", width = .15, height = .15), " ",
as_chunk(services, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
| -0.60% | 8.80% |
Y/Y |
|
|
|
| -1.10% | -0.70% |
Triennial |
|
|
|
| 4.60% | 4.60% |
Rolling 4Q |
|
|
|
| 2.50% | -3.60% |
dfxt2 <- flextable::compose(dfxt2,
j = 6,
value = as_paragraph(
as_image(src = "arrow_up.png", width = .15, height = .15), " ",
as_chunk(production, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
| 8.80% |
Y/Y |
|
|
|
|
| -0.70% |
Triennial |
|
|
|
|
| 4.60% |
Rolling 4Q |
|
|
|
|
| -3.60% |
dfxt2 <- flextable::compose(dfxt2,
j = 7,
value = as_paragraph(
as_image(src = "arrow_up.png", width = .15, height = .15), " ",
as_chunk(construction, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
|
|
Y/Y |
|
|
|
|
|
|
Triennial |
|
|
|
|
|
|
Rolling 4Q |
|
|
|
|
|
|
dfxt2<- flextable::compose(dfxt2,
j = 6, i = ~production < 0.011,
value = as_paragraph(
as_image(src = "arrow_down.png", width = .15, height = .15), " ",
as_chunk(production, props = fp_text(color = "black"))
),
part = "body")
dfxt2growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
|
|
Y/Y |
|
|
|
|
|
|
Triennial |
|
|
|
|
|
|
Rolling 4Q |
|
|
|
|
|
|
dfxt2 <- flextable::compose(dfxt2,
j = 7, i = ~construction < 0.04,
value = as_paragraph(
as_image(src = "arrow_down.png", width = .15, height = .15), " ",
as_chunk(construction, props = fp_text(color = "black"))
),
part = "body")
dfxt2 #this should be the completed body growth | NICEI | private | public | services | production | construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
|
|
Y/Y |
|
|
|
|
|
|
Triennial |
|
|
|
|
|
|
Rolling 4Q |
|
|
|
|
|
|
Step 4. Final prettification
The arrows are all the correct colour and attributed to the correct values. Now the column names can be changed back to the correct format.
dfxt3 <- set_header_labels(dfxt2,
values = list(
growth = "Growth Rates",
NICEI = "NICEI",
private = "Private Sector",
public = "Public Sector",
services = "Services",
production = "Production",
construction = "Construction"
)
)
dfxt3Growth Rates | NICEI | Private Sector | Public Sector | Services | Production | Construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
|
|
Y/Y |
|
|
|
|
|
|
Triennial |
|
|
|
|
|
|
Rolling 4Q |
|
|
|
|
|
|
Finally, the last act is to amend the column widths and text alignment to make the table more aesthetically pleasing and readable
#prettifying the table via formatting to reflect the publication be replicated
dfxt4 <- width(dfxt3, j = 1, width = 1.3086) %>% # Adjusting the width of column 1
align(align = "center", part = "all") %>% # Center aligning all of table contents
align(j=1, align = "left", part = "body") # Left aligning column 1
growth_table <-dfxt4 # Final table
growth_tableGrowth Rates | NICEI | Private Sector | Public Sector | Services | Production | Construction |
|---|---|---|---|---|---|---|
Q/Q |
|
|
|
|
|
|
Y/Y |
|
|
|
|
|
|
Triennial |
|
|
|
|
|
|
Rolling 4Q |
|
|
|
|
|
|
3. Contribution to Quarterly Change in NICEI (Q4 2022)
Step 1: Installing relevant libraries
library(readxl)
library(tidyverse)
library(formattable)
library(ggplot2)Step 2: Data Import and Preparation
Loading data
Import from Excel spreadsheet in local directory
df <- read_excel("NICEI-Tables-Q4-2022.xlsx",
sheet = "Table 6",
range = "A2:B8")Data preparation
In the publication, the plot has an evident gap between the NICEI bar and the rest. To replicate this in R, an empty row will be created plotted. Row 5 does not contain any values, only NAs, which for the characters column (1) will need placed with an empty value.
# Replace row 5 column 1 NA with empty space.
df[5,1]<-""
formattable(df) #table check| Sector | Quarterly contribution (pps)* |
|---|---|
| Services | 0.5 |
| Production | -0.1 |
| Construction | 0.6 |
| Public Sector | 0.1 |
| NA | |
| NICEI (% change) | 1.4 |
Step 3. Plotting
Initial plot
Initial plot of histogram to check layout and comparison to original
ggplot(df, # data
aes(Sector, # X values
`Quarterly contribution (pps)*`)) + # Y values
geom_col() + # Plot type
coord_flip() # Horizontal layout
The plot needs quite some work. The order is incorrect, the bars are too thick, there is no colour, no data labels and there are both axis titles and background colour.
Plot preparation
This code chunk attempts to ensure the correct order of sequence and generate the appropriate colour scheme. Note there are only 3 colours but 5 variables. A group will need to be created and according to colour, attributed the variables.
df1<-map_df(df, rev) # reverse the order of the columns in df
df1$category<- c("A", "B", "B", "B", "B", "C") # creating a new column to assist with colour scheme in original
formattable(df1) # table check| Sector | Quarterly contribution (pps)* | category |
|---|---|---|
| NICEI (% change) | 1.4 | A |
| NA | B | |
| Public Sector | 0.1 | B |
| Construction | 0.6 | B |
| Production | -0.1 | B |
| Services | 0.5 | C |
The table is now in the order required for plotting and the categories can be used to allocated colour to the bars. Time to plot check.
hbplt<-ggplot(df1,
aes(Sector, # X-axis values
`Quarterly contribution (pps)*`, # Y-axis
fill = category)) + # colours based on column "category"
geom_col() + # plot type
coord_flip() # flip plot to horizontal
hbplt
The colours are allocated correctly but the sequence is still wrong; it isn’t plotting in same order as table. To ensure fixed sequencing another column will be created called “order” which will number the rows 1:5 based on the sequence they should be plotted.
df2<-df1 # creating new data frame based on previous
df2$order<-c(1:nrow(df2)) # adding row attributing an order for the plot
hbplt2 <- ggplot(df2,
aes(fct_reorder(Sector, desc(order)), # allocates Sector plotting sequence based on 'order'
`Quarterly contribution (pps)*`,
fill = category)) +
geom_col(width = 0.25) + # changes column width so they're thinner
coord_flip()
hbplt2
The order column works, now to finish the plotting aesthetics
Final Plotting
final_plot <- hbplt2 +
theme_classic() + # blank background as per original image
geom_text(aes(label = # adding data labels to the plot
ifelse(`Quarterly contribution (pps)*` < 0, `Quarterly contribution (pps)*`, "")), # if value is negative
hjust = 1.1) + # position label on left
geom_text(aes(label =
ifelse( `Quarterly contribution (pps)*` > 0, `Quarterly contribution (pps)*`, "")), # if value is negative
hjust = -0.2) + # position label on left
scale_fill_manual(values= c("#000066", # colour first category
"blue", # colour second category
"#cccc00")) + # colour third category
theme(aspect.ratio = 1/2, # change aspect ratio of plot
plot.margin = unit(c(2,2,2,2), "cm"), # pad area around plot for image capture
panel.grid.major.x = element_line(color = "grey"), # colour major gridlines grey
plot.title = element_text(hjust = 0.5, # centre the plot title and bolden
face="bold"),
axis.text = element_text(size=10, face = # make axis text bold and size 10
"bold"),
axis.line = element_line(colour = "grey"), # colour axis lines grey
axis.ticks = element_blank(), # remove axis ticks
legend.position="none") + # no legend
scale_y_continuous(limits=c(-0.5,1.5), # alter y axis to start at -0.5 and end at 1.5
expand = c(0, 0)) + # y axis limits to start at plot origin
labs(x = "", # plot to have no X- or Y- titles but a main plot title
y="",
title = "Contributions of component indices to quarterly change in \nNICEI** (pps)"
)
final_plot #sanity check
Step 4. (Optional) Script to save plot as image
As before, the plot image can be manually exported or a code chunk can be inserted to automtically do this; however, when this is executed the text sizes are much smaller and can be difficult to read. to fix this the code chunk from above is repeated but with the text sizes need amended.
Repeat code chunk but amend text sizes
final_plot_sized <- ggplot(df2,
aes(fct_reorder(Sector, desc(order)), # attributes a sequence of plotting Sector based on 'order'
`Quarterly contribution (pps)*`,
fill = category)) +
geom_col(width = 0.25) + # adjusts the column widths so they are thinner
theme_classic() + # blank background as per original image
geom_text(aes(label = # adding data labels to the plot
ifelse(`Quarterly contribution (pps)*` < 0, `Quarterly contribution (pps)*`, "")), # if value is negative
hjust = 1.1) + # position label on left
geom_text(aes(label =
ifelse( `Quarterly contribution (pps)*` > 0, `Quarterly contribution (pps)*`, "")), # if value is negative
hjust = -0.2) + # position label on left
scale_fill_manual(values= c("#000066", # colour first category
"blue", # colour second category
"#cccc00")) + # colour third category
theme(aspect.ratio = 1/2, # change aspect ratio of plot
plot.margin = unit(c(2,2,2,2), "cm"), # pad area around plot for image capture
panel.grid.major.x = element_line(color = "grey"), # colour major gridlines grey
plot.title = element_text(size = 17, # centre the plot title, size = 17 and bolden
hjust = 0.5,
face="bold"),
axis.text = element_text(size=13, face = # make axis text bold and size 13
"bold"),
axis.line = element_line(colour = "grey"), # colour axis lines grey
axis.ticks = element_blank(), # remove axis ticks
legend.position="none") + # no legend
coord_flip()+ # flip plot to horizontal
scale_y_continuous(limits=c(-0.5,1.5), # alter y axis to start at -0.5 and end at 1.5
expand = c(0, 0)) + # y axis limits to start at plot origin
labs(x = "", # plot to have no X- or Y- titles but a main plot title
y="",
title = "Contributions of component indices to quarterly change in \nNICEI** (pps)"
)Automatic plot image export
png("contributions_resized.png", width = 950, height = 750)
final_plot_sized
dev.off()